Determining missing data with similarity based impitation method and comparison with simple imputations method

In the research, it is aimed to develop similarity-based assignment method as a new method in completing missing observations and to compare this new method and descriptive statistics methods in terms of reliability and correct assignment rates. The research was designed in the context of the basic research model. The study group of the research consists of the data of 450 students selected by random numbers table among the 3rd grade students who answered the interest scale in the field of educational sciences at Akdeniz University. Cronbach's alpha reliability test and comparison analyzes of correct assignment rates and descriptive analyzes on the data obtained in the study were performed with the spss package program. With the analyzes made, it was seen that the reliability coefficient of the similarity-based assignment method was higher than the descriptive statistics methods. In addition, it has been determined that the correct assignment rate in completing the missing observation with the similarity-based assignment method is higher than the descriptive statistics methods. According to the findings, it was concluded that the most accurate assignment method was the similarity-based assignment method (5%).